Table 3. Associations of dietary GL, GI, and total carbohydrate intake with a poor functional outcome.

Variables	GL	GI	Total carbohydrates
Binary logistic regression (mRS score=0–2 versus 3–6)
Continuous variables, OR (95% CI), p	1.01 (1.00–1.02), 0.003	1.00 (0.89–1.12), 0.982	1.01 (1.00–1.02), 0.005
Ordinal variables, quartile; OR (95% CI), p for trend	3.38 (1.66–6.90), 0.002	1.20 (0.71–2.03), 0.481	3.08 (1.50–6.32), 0.002
Q1, Reference	1	1	1
Q2, OR (95% CI), p	6.31 (0.59–67.39), 0.127	4.25 (0.69–25.94), 0.117	12.64 (1.12–142.14), 0.040
Q3, OR (95% CI), p	32.78 (2.81–381.86), 0.005	1.83 (0.26–12.63), 0.538	20.64 (1.68–253.56), 0.018
Q4, OR (95% CI), p	28.93 (2.82–296.04), 0.005	3.60 (0.54–23.76), 0.182	36.84 (2.99–453.42), 0.005
Ordinal logistic regression (mRS score=0, 1, 2, 3–6)
Continuous variables, OR (95% CI), p	1.00 (1.00–1.03), 0.005	1.03 (0.95–1.11), 0.519	1.00 (1.00–1.03), 0.006
Ordinal variables, quartile; OR (95% CI), p for trend	1.58 (1.10–2.27), 0.014	1.20 (0.86–1.69), 0.284	1.69 (1.16–2.45), 0.006
Q1, Reference	1	1	1
Q2, OR (95% CI), p	1.06 (0.51–6.63), 0.356	1.99 (0.67–5.91), 0.215	2.67 (0.73–9.72), 0.137
Q3, OR (95% CI), p	6.82 (1.99–23.57), 0.002	1.77 (0.57–5.46), 0.321	7.66 (2.08–28.22), 0.002
Q4, OR (95% CI), p	4.71 (1.32–16.78), 0.017	2.02 (0.66–6.20), 0.217	5.00 (1.35–18.62), 0.016

Adjusted for age, sex, variables with p<0.1 in univariate analysis (marital status, prestroke mRS score, hyperlipidemia, body mass index, stroke classification, National Institutes of Health Stroke Scale score, early neurological deterioration, and high-grade white-matter hyperintensities), and variables related to GL, GI, or total carbohydrate intake (diabetes mellitus, triglycerides, low-density lipoprotein, and hemoglobin A1c).

CI: confidence interval, GI: glycemic index, GL: glycemic load, mRS: modified Rankin Scale, OR: odds ratio.